For the last few months, a team of geography students at SDSU have been working with the crime data provided by the Library, producing analyses and visualizations of the data.

Elias Issa has been looking at Drugs and Alcohol violations in Downtown San Diego and East Village. He writes:

The Hot Spot tool calculates the  Gi* statistic for each feature in a dataset. The resultant z-scores and p-values tell you where features with either high or low values cluster spatially.  To have a statistically significant hot spot, a feature will have a high value and be surrounded by other features with high values as well. My animated maps illustrate 2 hot spots from 2007 to 2012. the most significant hot spot is located close to St. Vincent Paul homeless shelter ( Imperial Ave) on the South Eastern part of East Village. The second hot spot is located North of Gaslamp Quarter between Broadway And Market which most of the famous and popular bars  are found. In addition to those map,and based on those hot spots, I did some statistical analysis to show the average of Drugs/ Alcohol violation monthly (Above/ Below Avg) and yearly. My study reveals that there is a slight increase within 2011 and 2012 in the average of Drugs/ Alcohol violations.

Over the course of the project, he has been experimenting with various ways to visualize the time component of geographic data. This is quite difficult, since you can’t easily scan the time dimension like you can in space. Visual processing is tuned for noticing changes and differences — like a deer that won’t notice you if you don’t move — so Elias’ visualization is best for quickly identifying areas that deserve more analysis, rather than showing the quantitative differences.

Due to the difficulty of  creating animations like this in ArcMap, the video has only one frame per year, but that is enough to illustrate how the changes from frame to frame draw your eye to problem areas. Without a visualization like this, it is easy to miss some of the most important features of an issue including short term spikes and long-term trends.

After identifying an area to focus on through the visualization, Elias’ underlying statistical method serves to quantify  the differences between times or locations, so this project is a great example of a way to use animation to partition a large problem space into components that can be analyzed in detail.